Catalysis Today 279 (2017) 194–201

Contents lists available at ScienceDirect

Catalysis Today
journal homepage: www.elsevier.com/locate/cattod

The influence of metal selection on catalyst activity for the liquid

phase hydrogenation of furfural to furfuryl alcohol
Á. O’Driscoll a,b,∗ , J.J. Leahy a,b , T. Curtin a,b
a
Carbolea Research Group, Department of Chemical and Environmental Sciences, University of Limerick, Limerick, Ireland
b
Materials and Surface Science Institute (MSSI), University of Limerick, Limerick, Ireland

a r t i c l e i n f o a b s t r a c t

Article history: In this work the replacement of toxic chromium containing catalysts for the selective hydrogenation
Received 14 December 2015 of furfural to furfuryl alcohol was investigated. The initial focus was on the synthesis of monometal-
Received in revised form 2 June 2016 lic catalysts by wet impregnation and concentrated on the employment of metals such as platinum,
Accepted 4 June 2016
palladium, copper and nickel. Experiments were conducted using ethanol as the solvent which was
Available online 18 June 2016
found to have a negative effect on the selectivity to the desired product, furfuryl alcohol, with high
quantities of 2-Furaldehyde diethyl acetal and difurfuryl ether formed. Consequently, toluene was
Keywords:
selected as an alternative solvent facilitating selectivity to furfuryl alcohol only. It was found that
Furfural
Furfuryl alcohol
platinum was the most promising metal of those studied as it displayed higher selectivity to furfuryl
Platinum catalyst alcohol and was subsequently employed for the synthesis of bimetallic catalysts. The bimetallic cata-
Surface organometallic catalysis lysts were synthesised by surface reactions using a variety of promoter metals selected according on
Liquid phase hydrogenation their electronegativity. It was found that, while the selectivity of all catalysts to furfuryl alcohol was
Biomass close to 100%, the conversion was influenced significantly by the second metal and followed the order
tin > molybdenum > manganese > barium > iron > nickel. The purpose of the research was to produce an
active catalyst for the liquid phase hydrogenation under suitable industrial conditions with the results
presented here conducted at 100 ◦ C and 20 bar hydrogen pressure. Furfural conversion of 47% and close
to 100% selectivity to furfuryl alcohol was achieved using a 0.6%Pt0.4%Sn/SiO2 catalyst.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction Furfural, produced by the acid-catalysed dehydration or hydrol-

Research to develop alternative liquid transport fuels has
increased considerably in recent years influenced by the imple-
mentation of legislation and directives on alternative energy which
sets targets for transposing members. Europe has set targets such as
the Renewable Energy Directive to make renewable energy sources
account for 20% of overall energy and 10% of transport energy
by 2020 [1]. To achieve these targets in an ethical, environmen-
tally sustainable and economical manner the use of lignocellulosic
biomass or agri-waste for the production of second generation
biofuels is proposed [2]. The production of biofuels from hemi-
cellulose requires acid hydrolysis to open the biomass structure
giving large quantities of xylose [3]. The synthesis of furfural (FF)
as a by-product of this biofuel production process has developed
an additional important research area.
∗ Corresponding author at: Carbolea Research Group, Department of Chemical
and Environmental Sciences, University of Limerick, Limerick, Ireland.
E-mail address: aine.odriscoll@ul.ie (Á. O’Driscoll).
ysis of xylose, is a versatile and renewable chemical with wide
industrial applications [4–6]. These include uses in the agrochem-
ical, fragrance and plastics industries while it is also a building
block for the synthesis of various other chemicals and chemical
products including nylons and adhesives [5,7,8]. The hydrogena-
tion of furfural to furfuryl alcohol (FA) is a single step process and
the most popular application of furfural conversion with approx-
imately 60–62% of the furfural produced worldwide converted to
furfuryl alcohol [9].
Furfuryl alcohol is formed from a simple hydrogenation reaction
in the presence of a catalyst. The C O of the aldehyde functional
group in furfural breaks while the addition of hydrogen to the oxy-
gen results in an alcohol group (O H) with an additional hydrogen
bonded to the exocyclic carbon of the furan based structure giv-
ing furfuryl alcohol. Furfuryl alcohol is used mainly in the polymer
industry together with the production of synthetic fibres, rubbers,
resins and farm chemicals. It is used in the manufacture of lysine,
vitamin C and lubricants, the production of foundry sand binders
in the metal casting industry and as a chemical building block for

http://dx.doi.org/10.1016/j.cattod.2016.06.013
0920-5861/© 2016 Elsevier B.V. All rights reserved.
 Á. O’Driscoll et al. / Catalysis Today 279 (2017) 194–201
Fig. 1. The hydrogenation pathway of the platform chemical, furfural.
drug synthesis. Furfuryl alcohol is also used in the production of
other products in the fine chemical industry [10–13].
The derivatives of furfural, some of which are outlined in Fig. 1,
include furfuryl alcohol, tetrahydrofuran, 2-methylfuran, furoic
acid, hydroxymethylfurfural, furylidene, furan and resins [14–17].
Many of these products are formed via furfural hydrogenation with
a variety of additional reactions including furfural decarbonylation
and oxidation also possible. It is therefore of great importance to
choose a catalyst which is selective to the desired reaction route.
The industrial catalyst for the hydrogenation of furfural to
furfuryl alcohol is copper chromite and is well known to be environ-
mentally toxic, promoting extensive research to develop suitable
alternatives [18]. The necessity to produce an alternative has
increased the interest in heterogeneous catalysts for the synthesis
of fine chemicals.
A wide variety of chromium free catalysts exist in literature
employing an extensive selection of support and metal com-
binations. Catalysts investigated for liquid phase research were
examined as a basis for metal selections applied in this work. Met-
als frequently selected include copper [19–22] and nickel [23–26]
while metals such as platinum [27–29] and palladium [30–32] are
sparsely reported with publications limited to dedicated research
groups. Overall, no trends between metal selection and catalyst
activity are evident from the literature although bimetallic cata-
lysts are generally more active than monometallic catalysts for the
desired reaction.
Furfural may be hydrogenated to furfuryl alcohol in the liquid
or gas phase with the industrial process utilising high temper-
ature and pressure although the chromium containing catalyst
used in industry achieves only moderate activity for this reaction
[29,33,34]. Vapour phase is often favoured over liquid phase due
to the uneconomic outcome of large-scale batch processes. How-
ever, vapour phase also has issues relating to the production of
undesired derivatives at high furfural conversion together with
deactivation of the catalyst [35]. A wide variety of catalyst synthesis
techniques have been investigated for this reaction from tech-
niques such as impregnation [30,34] and co-precipitation [19,26] to
195
complex techniques including surface reactions together with syn-
thesising supported and encapsulated metal nanoparticle catalysts
[29,36–38].
This work focuses on the synthesis of a catalyst that is active for
the hydrogenation of furfural to furfuryl alcohol. The selectivity to
furfuryl alcohol is crucial which is the emphasis for the selection
of the primary metal. An analytical approach regarding the elec-
tronegativity of metals is applied to the selection of a second metal
in order to increase furfural conversion. The solvent used for the
reaction is also scrutinised to ensure the selection does not have a
negative effect on the desired reaction.
2. Materials and methods
Furfural, furfuryl alcohol, ethanol, toluene, hydrofluoric acid,
hexanol, 1-propanol and n-heptane, all of analytical reagent quality,
were obtained from Sigma-Aldrich. Metal salts for catalyst prepa-
ration; copper (II) nitrate hydrate, nickel (II) nitrate hexahydrate,
platinum (II) acetylacetonate, tetraaminepalladium (II) chloride
monohydrate, tetrabutyl tin, iron (II) chloride tetrahydrate, man-
ganese (II) acetate tetrahydrate, nickel (II) nitrate heptahydrate,
barium acetate molybdenum acetate together with lanthanum
(III) chloride heptahydrate, aluminium isopropoxide, pleuronic® P-
123 (P123) and tetraethyl orthosilicate (TEOS) were also obtained
from Sigma-Aldrich. Metal standards were obtained from Lennox
Dublin. Commercial palladium catalysts and catalyst supports, SiO2
and TiO2 were obtained from Sigma-Aldrich while ß-Zeolite was
obtained from Zeolyst International.
3. Experimental
3.1. Catalyst preparation
All monometallic catalysts were produced using wet impregna-
tion. The required amount of a precursor salt was added to 50 ml
of ethanol followed by the addition of the support. The suspen-
196 Á. O’Driscoll et al. / Catalysis Today 279 (2017) 194–201

Table 1
Catalysts tested for furfural conversion and furfuryl alcohol selectivity. (Reaction conditions: 1 g catalyst, 25 ml FF, 175 ml solvent (ethanol), PH2 = 20 bar, T = 100 ◦ C, t = 300 min).

Catalyst Expected Metal Loading (%) Measured Metal Loading (%) Conversion of Furfural (%) Selectivity to Furfuryl Alcohol (%)

Cu/SiO2 1 1.1 29 <1

Ni/SiO2 1 0.9 37 <1
Pd/SiO2 1 1.0 44 10
Pt/SiO2 1 0.6 31 29
Cu/SiO2 2 2.0 27 <1
Ni/SiO2 2 1.9 29 5
Pd/SiO2 2 1.9 45 8
Pt/SiO2 2 1.4 41 35

sion was stirred for 16 h at room temperature, followed by solvent

evaporation and calcination in air at 450 ◦ C for 300 min. nsFF = moles of furfural in sample
The catalyst support Al-SBA-15 was synthesized according to a
procedure reported by Srivastava et al. [39].
Surface reactions were used for the production of bimetallic cat- nFA = moles of furfuryl alcohol
alysts. A calcined platinum catalyst, produced by wet impregnation,
was placed in an autoclave (Parr 4560) and reduced at 300 ◦ C for
four hours at 20 bar hydrogen pressure. A solution was prepared
3.4. Catalyst characterisation
containing the required amount of the desired metal precursor
dissolved in 200 ml of solvent selected depending on the solu-
The catalysts were characterised by Atomic Absorption Spec-
bility of the salt. Iron, molybdenum, manganese and nickel salts
troscopy (AAS). Prior to AA analysis, digestion was performed using
were dissolved in 1-Propanol with tin and barium salts dissolved
an Anton Parr 8NXF100 autoclave where 25.0 mg catalyst was
in n-heptane and deionised water respectively. This solution was
added to 10 ml hydrochloric acid. The sample was digested with
degassed using nitrogen and added to the reactor and stirred at
a thermal input of 900 W using a three step method including
600 rpm for four hours at 90 ◦ C and 20 bar hydrogen pressure. This
15 min ramp, 60 min hold and 30 min cooling. Lanthanum chloride
suspension was then filtered and dried overnight at 60 ◦ C.
was added to the digested sample to eliminate interferences from
hydrochloric acid. A Varian Spectra 220 Atomic Absorption Spec-
3.2. Catalytic tests
trometer was used to analyse the samples. For analysis of platinum
catalysts an acetylene air flame and a platinum hollow cathode
The liquid phase hydrogenation of furfural was carried out in a
lamps were used at a wavelength of 265.9 nm with a slit width
600 ml autoclave (Parr 4560) equipped with a mechanical stirrer.
of 0.2 nm.
For a typical experiment, 1.0 g of catalyst was placed in the reactor
Scanning electron microscopy (SEM) & energy-dispersive X-ray
and reduced at 300 ◦ C and 20 bar of hydrogen pressure for 4 h. The
analysis (EDAX) were also used for measurement of metal content
reactor was then flushed with nitrogen and charged with 25 ml
using a Hitachi SU-70 SEM operating at 20 kV equipped with an
of furfural and 175 ml solvent. The reactor was heated to 100 ◦ C
EDAX detector. Samples were placed directly onto carbon tape on
and the pressure was rapidly increased to 20 bar hydrogen with a
the specimen stub with no further pre-treatment performed.
stirring speed of 600 rpm. Homogeneous mixing was assumed and
Surface area analysis was used to establish surface area, pore
samples were taken at regular intervals for analysis by gas chro-
volume and pore radius of the catalysts by nitrogen physisorption
matography (Agilent Technologies 7820A) equipped with an FID
using a Quantachrome Autosorb AS-1. For each analysis, 10 mg of
detector. A Restek Stabilwax® 10623 30 m × 0.25 mm × 0.25 ␮m
sample was outgassed under vacuum at 150 ◦ C for 12 h to remove
column was used, with nitrogen as the carrier gas at a flow rate of
water and other atmospheric contaminants followed by analysis
12 ml min−1 and a split ratio of 1:25. The injector and detector were
consisting of measurement of adsorption and desorption of nitro-
operated at 300 ◦ C and 250 ◦ C, respectively. A selection of samples
gen gas on the sample conducted at −196 ◦ C.
were analysed using gas chromatography-mass spectroscopy (Agi-
lent Technologies 7890A) under the same conditions for product
identification purposes. 4. Results and discussion

Initial catalyst testing focused on four selected metals, plat-

3.3. Formulae
inum, palladium, nickel and copper supported on silica for furfural
conversion and selectivity to furfuryl alcohol. A complete list of
The conversion of furfural, selectivity to furfuryl alcohol and
the catalysts tested is outlined in Table 1 which summarises the
reaction yield were calculated as follows;
hydrogenation performance of the prepared catalysts. Furfural con-
noFF − nSFF 100 version after 5 h reaction ranged from 27 to 45% indicating that all
FFConversion (%) : × catalysts showed some activity for the reaction. The selectivities
noFF 1
to furfuryl alcohol ranged from <1 to 35%. The overall low furfuryl
alcohol selectivity was as a result of the production of undesirable
nFA 100 products. The selectivity to the desired product was particularly
FASelectivity (%) : ×
noFF − nSFF 1 low for the nickel and copper based catalysts. Although in gen-
eral the catalyst activity was poor for the desired reaction, the
nFA 100 apparent influence of each metal was clear with palladium and plat-
Yield (%) : × inum indicating a positive influence on conversion and selectivity
noFF 1
respectively. Therefore, these two metals were the focus of catalyst
development which concentrated on increasing furfural conversion
noFF = initial moles of furfural and furfuryl alcohol selectivity.
 Á. O’Driscoll et al. / Catalysis Today 279 (2017) 194–201 197

Table 2
Palladium supported catalysts examined.

Palladium Supported Catalysts Expected Metal Loading (%) Actual Metal Loading (%) Surface Area (m2 /g)

Pd/SiO2 2.0 1.9 566

Pd/TiO2 2.0 0.4 16
Pd/Al-SBA-15 2.0 0.9 538
Pd/ß-Zeolite 2.0 1.6 509
10%Pd/C-Sigma Aldrich – 10.0a 950b
1%Pd/C-Sigma Aldrich – 1.0a 760b
a
Manufacturer Analysis.
b
Characterisation by [40].

60 10
1.6% Pd/ Zeolite
9 0.9% Pd/Al-SBA-15
50 0.4% Pd/TiO 2
8

7
1.9% Pd/SiO 2
FF Conversion (%)

40

FA Selectivity (%)
6

30 5

4
20
3
1.6% Pd/ Zeolite
0.9% Pd/Al-SBA-15 2
10 0.4% Pd/TiO2
1
1.9% Pd/SiO2
0 0
0 60 120 180 240 300 0 60 120 180 240 300

Time (min) Time (min)

Fig. 2. The conversion of furfural using the indicated synthesised catalysts. (Reaction Fig. 3. Influence of the indicated catalysts on selectivity to furfuryl alcohol with
conditions: 1 g catalyst, 25 ml FF, 175 ml solvent (ethanol), PH2 = 20 bar, T = 100 ◦ C). reaction time. (Reaction conditions: 1 g catalyst, 25 ml FF, 175 ml solvent (ethanol),
PH2 = 20 bar, T = 100 ◦ C).

5. Palladium supported catalysts

From initial testing, palladium was observed to be the most
influential metal for furfural conversion. Therefore, this catalyst
was investigated further with the aim of increasing furfural con-
version and improving furfuryl alcohol selectivity.
A selection of supports were chosen and used in the synthesis
of a number of palladium catalysts. The supports chosen included
Al-SBA-15, ß-Zeolite and TiO2 . All catalysts were prepared in a sim-
ilar way (wet impregnation) with a view to investigation the role
of support. The hydrogenation performance of the synthesised cat-
alysts was compared to commercial palladium catalysts obtained
from Sigma Aldrich. A complete list of these catalysts is shown in
Table 2 which also presents the B.E.T surface areas of the synthe-
sised and commercial catalysts. The expected palladium loading for
the synthesised catalysts was 2 wt% and the table also includes the
atomic absorption measured palladium loading for each of these
prepared catalysts. The highest surface area was exhibited by the
commercial catalysts from Sigma-Aldrich, 1%Pd/C and 10%Pd/C,
which used activated carbon as the support. The synthesised cat-
alysts based on SiO2 , Al-SBA-15 and ß-Zeolite had comparable
surface areas however TiO2 displayed a much lower surface area.
The hydrogenation results for furfural conversion using the syn-
thesised catalysts 1.9%Pd/SiO2 , 0.4%Pd/TiO2 , 0.9%Pd/Al-SBA-15 and
1.6%Pd/ß-Zeolite are presented in Fig. 2. A difference in furfural
conversion was seen for the selection of supports studied indicat-
ing that the supports may influence the conversion. This is most
notable when 1.9%Pd/SiO2 and 1.6%Pd/ß-Zeolite are compared as
they have similar surface areas but despite the ß-Zeolite having
a lower palladium content it displays higher furfural conversion.
There are many factors which may influence the catalyst activity
including the acidity of the support, the interaction between the
metal and the support, the surface area and the metal content. It
was observed that low conversion occurred when surface area and
metal content were also low.
Fig. 3 shows the selectivity to furfuryl alcohol with hydrogena-
tion time for the prepared catalysts. Overall, the selectivity was
poor however it seems that selectivity to furfuryl alcohol was
promoted by using silica as a support. The furfural conversion
and furfuryl alcohol selectivity results may indicate the impor-
tance of the acidity of the supports on the overall catalyst activity.
The presence of aluminium in the beta zeolite and the Al-SBA-15
makes these supports acidic in nature and both of these catalysts
presented the highest furfural conversions. However, although
these acidic supports resulted in the promotion of furfural con-
version they presented low furfuryl alcohol selectivity. This may
be a result of furfural acetalization which is a non-catalytic reac-
tion facilitating the conversion of furfural due to the presence of
ethanol. Alternatively, the low furfuryl alcohol selectivity may be
due to etherification with ethanol as the Brønsted acid supports
can influence the direction of the reaction while Lewis acid sup-
ports such as TiO2 facilitate furfuryl alcohol etherification via the
Meerwein-Pondorf-Verley reaction whereby the carbonyl group
and the alcohol coordinate to a Lewis acid metal centre and a
hydride transfer from the alcohol to the carbonyl occurs [41–44].
TEM analysis was performed on the synthesised palladium cat-
alysts. Overall, the dispersion of the metal particles was similar
for all catalysts (Supplementary Information, Fig. A1–4) with metal
particles of 2–5 nm present on all supports with the exception of
0.4%Pd/TiO2, most likely due to the low metal loading. The relatively
large particle sizes observed were most likely due to the method of
preparation used.
Silica was selected as the support for all further synthesised
catalysts as it showed the highest selectivity to furfuryl alcohol.
Table 3 presents the activity of the commercial palladium cata-
lysts at 25 ◦ C and 100 ◦ C and compares them to the activity of the
synthesised 1.9%Pd/SiO2 . The reaction temperature seems to have
no clear influence on the reaction. High conversions were achieved
198 Á. O’Driscoll et al. / Catalysis Today 279 (2017) 194–201

Table 3
A comparison of synthesised and commercial palladium catalysts. (Reaction condi-
tions: 1 g catalyst, 25 ml FF, 175 ml solvent (ethanol),PH2 = 20 bar, t = 300 min).

Catalyst Temperature (◦ C) Conversion (%) Selectivity (%)

1.9%Pd/SiO2 (Synthesised) 100 45 8

10%Pd/C (Sigma) 100 98 3
10%Pd/C (Sigma) 25 86 12
1%Pd/C (Sigma) 25 37 2

Fig. 5. The influence of metal selection and solvent selection on furfural conversion.
The percentage selectivity to furfuryl alcohol is also included for each test. (Reaction
conditions: 1 g catalyst, 25 ml FF, 175 ml Solvent, PH2 = 20 bar, T = 100 ◦ C, t = 300 min).

Fig. 4. Products detected at 35% furfural conversion for the indicated catalysts.
(Reaction conditions: 1 g catalyst, 25 ml FF, 175 ml solvent (ethanol), PH2 = 20 bar,
T = 100 ◦ C).

when the palladium content of the catalyst was high (10 wt%). How-
ever, high conversions tended to lead to low product selectivity.

5.1. Identification of additional products

A more detailed study was carried out to investigate the rea-

sons for the relatively low selectivity to furfuryl alcohol formation.
The products formed following hydrogenation were analysed using
GC–MS. Two additional products were identified and these were
observed for all hydrogenations conducted in ethanol. These were
difurfuryl ether and 2-Furaldehyde diethyl acetal. Indeed, the
selectivity to the desired product furfuryl alcohol was low in
the presence of ethanol. This was particularly true for the syn-
thesised catalysts and can be seen in Fig. 4 which presents the
reaction products of two synthesised catalysts 1.9%Pd/SiO2 and
1.4%Pt/SiO2 at 35% furfural conversion. These two catalysts were
selected as, of all the catalysts prepared, they were the best in
terms of furfural conversion and furfuryl alcohol selectivity respec-
tively, as outlined in Table 1. The measurement in Fig. 4 is given in
counts as difurfuryl ether is not commercially available prevent-
ing the preparation of a standard plot for calibration purposes. The
product formed using the synthesised catalysts was primarily 2-
furaldehyde diethyl acetal. This was followed by furfuryl alcohol
and difurfuryl ether. The formation of the desired product furfuryl
alcohol was low for both catalysts.
In order to investigate the influence of the solvent on furfural
conversion, 1.9%Pd/SiO2 and 1.4%Pt/SiO2 were hydrogenated using
ethanol (EtOH) and toluene (Tol) as solvents. The results of these
tests are displayed in Fig. 5, which shows the conversion of fur-
fural after 300 min hydrogenation and the selectivity to the desired
product, furfuryl alcohol (values given in brackets). From this study,
it was observed that furfural conversion is higher in the pres-
ence of ethanol with a slightly higher conversion observed using
a palladium catalyst. Although conversion is lower when toluene
is employed no difference in activity was observed between the
two catalysts tested. Selectivity to furfuryl alcohol varied in ethanol
Fig. 6. The influence of solvent on furfural conversion and product formation.
(Reaction conditions: no catalyst/1 g catalyst, 25 ml FF, 175 ml Solvent, PH2 = 20 bar,
T = 100 ◦ C, t = 300 min).
with platinum giving a selectivity value (35%) over four times that
of palladium (8%). Meanwhile, furfuryl alcohol was the principal
product formed when toluene was used as solvent (99% selectivity
for both catalysts). From this study it would seem that the platinum
based catalyst is slightly better in terms of selectivity to the desired
product, particularly in the presence of the ethanol solvent. There-
fore this catalyst was studied in more detail. It is also clear from
this study that the solvent has a major influence on the selectivity
to the various products, particularly the desired product, furfuryl
alcohol.
In order to further investigate the influence of the solvent, fur-
fural hydrogenation tests were carried out in the absence and in
the presence of the 1.4%Pt/SiO2 catalyst using ethanol and toluene
as solvent. The 1.4%Pt/SiO2 catalyst was chosen as it was the most
selective to furfuryl alcohol of the initial catalysts tested. The results
from this study are displayed in Fig. 6 which shows the percent-
age conversion of furfural and the formation of both desirable and
undesirable products (given in counts), in the absence and pres-
ence of the catalyst. Two solvents were investigated, ethanol and
toluene. It was observed that in the absence of a catalyst conver-
sion was very low using toluene as the solvent coupled with low
product formation. Furfural conversion was significantly higher,
 Á. O’Driscoll et al. / Catalysis Today 279 (2017) 194–201 199

Fig. 7. The reaction pathway for the acetalization of furfural to 2-furaldehyde diethyl acetal.
Source: adapted from Rubio-Caballero et al. [45].

Fig. 8. The reaction pathway for the etherification of furfuryl alcohol to difurfuryl ether.
Source: adapted from Zhang et al. [47].

in the absence of a catalyst, using ethanol as the solvent with 2- Table 4

furaldehyde diethyl acetal observed as the primary product. This
indicated that a reaction occurred between ethanol and furfural in
the homogeneous phase under the conditions studied. When a cat-
alyst was used, the conversion was higher for both solvents. It was
observed that using ethanol, furfuryl alcohol and 2-furfuraldehyde
diethyl acetal were formed while the use of toluene had a lower fur-
fural conversion but furfuryl alcohol was the main product formed
with minimal quantities of other by-products observed.
The results shown in Fig. 6 clearly indicate that 2-Furaldehyde
diethyl acetal is only formed in large quantities when ethanol is
present. In the presence of toluene, 2-furaldehyde diethyl acetal
was absent. The production of furfuryl alcohol was higher when a
catalyst was used with results indicating that a catalyst is essential
for the desired reaction to occur. The production of furfuryl alcohol
is higher when ethanol is employed, most likely resulting from the
higher furfural conversion observed with the use of ethanol. The
production of difurfuryl ether was only observed in the presence
of a catalyst using ethanol as solvent. Generally, the use of toluene
as a solvent resulted in the production of furfuryl alcohol only but
conversion was higher when ethanol was used.
In Table 1, the catalysts 0.9%Ni/SiO2 , 1.1%Cu/SiO2 , 1.9%Ni/SiO2
and 2% Cu/SiO2 all presented relatively similar furfural conversion
(∼30%) in the presence of ethanol solvent. From Fig. 6 it can be
seen that in the absence of a catalyst, 26% conversion is achieved.
This would indicate that the nickel and copper based catalysts are
not active for furfural conversion under the conditions studied and
the conversions observed in Table 1 are primarily the result of the
liquid phase non-catalytic acetalization reaction.
The proposed reaction pathway for the formation of 2-
furaldehyde diethyl acetal is presented in Fig. 7 while that for
difurfuryl ether is presented in Fig. 8. The carbonyl group of fur-
fural (1) is protonated from a H+ ion sourced from ethanol or
a catalyst to form an intermediate (2). Ethanol reacts with this
intermediate forming the hemiacetal. A proton is removed from
the hemiacetal which subsequently re-protonates and dehydrates
Prepared platinum supported on silica catalysts.
Catalyst Expected Pt Weight Loading (%) Measured Pt Weight Loading (%)
Pt/SiO2 0.5 0.3
Pt/SiO2 1 0.9
Pt/SiO2 2 1.9
to form another intermediate (4). This intermediate then reacts
with another ethanol molecule giving an intermediate (5) which is
deprotonated to form 2-furaldehyde diethyl acetal (6) [45,46]. The
etherification of furfuryl alcohol pathway shown in Fig. 8 begins
with the protonation of furfuryl alcohol (1), where it is proposed
that the H+ ion is available from the ethanol or a catalyst. This step
results in a protonated intermediate (2) which is dehydrated to
form an additional intermediate (3) that in turn reacts with another
furfuryl alcohol molecule to form difurfuryl ether (4) [47–49]. In
this work, these products were observed in the absence of a cat-
alyst therefore the H+ ion is most likely sourced from the solvent
ethanol rather than the catalyst.
6. Development of platinum catalysts
From the initial catalysts tested, platinum was observed as the
superior metal for furfuryl alcohol production. The development of
palladium catalysts resulted in relatively low furfuryl alcohol selec-
tivities despite the increase in furfural conversion therefore it was
decided to investigate the platinum based catalysts in more detail.
The influence of platinum metal loading on furfural conversion was
investigated by synthesising a series of platinum catalysts by wet
impregnation using silica as the support, these catalysts are listed
in Table 4.
Fig. 9 shows the percentage furfural conversion for the prepared
platinum supported silica catalysts with reaction time. The results
show that furfural conversion for 0.9%Pt/SiO2 and 1.9%Pt/SiO2 are
relatively similar with no difference after 300 min reaction time
200 Á. O’Driscoll et al. / Catalysis Today 279 (2017) 194–201

Table 5
Catalysts produced by surface organometallic catalysis.

Metal Electronegativity (Pauling) Actual Metal Loading Actual Platinum Loading

Platinum (Pt)- Base 2.28 – 0.9

Tin (Sn) 1.96 0.4 0.6
Molybdenum (Mo) 2.16 0.2 0.6
Manganese (Mn) 1.55 0.3 0.6
Barium(Ba) 0.89 0.2 0.6
Iron (Fe) 1.83 0.2 0.6
Nickel (Ni) 1.91 0.5 0.6

16 50
0.3% Pt/SiO2 Pt
14
Pt-Sn
0.9% Pt/SiO2
Pt-Mo
1.9% Pt/SiO2 40
Pt-Mn
12
Pt-Ba

FF Conversion (%)
FF Conversion (%)

Pt-Fe
10 30 Pt-Ni
8

20
6

4
10
2

0 0
60 120 180 240 300 60 120 180 240 300

Time (min) Time (min)

Fig. 9. The influence of metal loading on furfural conversion. (Reaction conditions: Fig. 10. The influence of a variety of promoters on the monometallic platinum
1 g catalyst, 25 ml FF, 175 ml solvent (toluene), PH2 = 20 bar, T = 100 ◦ C). base catalyst. (Reaction conditions: 1 g catalyst, 25 ml FF, 175 ml solvent (toluene),
PH2 = 20 bar, T=100 ◦ C).

and considerably better than 0.3%Pt/SiO2 . This result is most likely
due to the dispersion of platinum on the support. As the platinum
loading increased it is possible that metal agglomeration occurred
resulting in low platinum dispersion. It is well known that low
metal dispersions can lead to poor catalytic activities [50]. Selec-
tivity to furfuryl alcohol was close to 100% for all catalysts. Thus,
after considering all factors, a Pt/SiO2 catalyst with a desired 1 wt%
loading was selected for future work involving the investigation of
using bimetallic catalysts.
The research into bimetallic catalysts focused on the synthe-
sis of catalysts by controlled surface reactions as the method was
observed in literature to produce catalysts with low metal loadings
together with good activity and was therefore suited to industrial
requirements. A variety of metals were selected to act as pro-
moters of the selected base catalyst, 1%Pt/SiO2 . This assortment
included tin, molybdenum, manganese, barium, iron and nickel.
These metals were selected according to published trends as Tathod
and Dhepe [51] state that the addition of promoter metals which
have an electron deficient state in the bimetallic catalyst are supe-
rior hydrogenation catalysts as they polarise the furfural carbonyl
group. Meanwhile, Merlo et al. [28] suggested that the difference
in electronegativity of the metals in the bimetallic system also
influences the activity of the catalyst. A promoter metal which is
more electropositive than the base metal is deemed the best as
it increases interaction with the furfural carbonyl group. Research
undertaken by the group was directed at increasing the selectiv-
ity of the reaction to furfuryl alcohol. However, the principle was
applied to this work with a view to increasing furfural conversion.
The electronegativities of the selected metals are shown in Table 5
which illustrates that a varied range of metals which are more elec-
tropositive than platinum were selected. A more electronegative
metal than platinum was not selected for comparison as platinum
is one of the most electronegative metals.
A complete list of the catalysts examined is shown in Table 5. The
expected metal loadings for the promoter metal and the platinum
were 0.4% and 1% respectively.
All catalysts were tested for the hydrogenation of furfural to
furfuryl alcohol using the same reaction conditions. Fig. 10 shows
the conversion of furfural with reaction time for all the prepared
bimetallic catalysts and these are compared to the monometallic
platinum catalyst. The platinum loading of these bimetallic cat-
alysts was similar as all were synthesised by wet-impregnation
from a single batch. However, some variation was observed for
the loading of the promoter metals which were deposited by sur-
face reactions. These variations are most likely due to issues arising
from the stirred metal solution used in the catalyst synthesis. This
includes low solubility of the metal salt and insufficient degassing
causing reduced metal deposition.
The monometallic platinum catalyst achieved 14% furfural con-
version after 5 h. Overall, the addition of a second metal was seen
to have a positive influence on the desired reaction with the excep-
tion of nickel. Several of the promoter metals studied resulted in
a significant increase in furfural conversion including manganese
and molybdenum, obtaining 33% and 38% conversion respectively.
However, the bimetallic catalysts containing Pt-Ni, Pt-Fe and Pt-Ba
yielded furfural conversion similar to the monometallic platinum
catalyst. Tin was found to have the most significant influence by
increasing the furfural conversion to 47%.For the conditions stud-
ied, selectivity to furfuryl alcohol did not vary across the metal
selection, remaining close to 100% for all prepared catalysts.
The influence of each promoter metal studied was clearly
observed however no correlation was found between the elec-
 Á. O’Driscoll et al. / Catalysis Today 279 (2017) 194–201
tronegativities of the metals investigated and their influence on
the furfural conversion. The difference in electronegativity of pro-
moter metals was previously studied [29] to investigate furfuryl
alcohol selectivity and side-product formation. A comparable influ-
ence on the furfural conversion as studied in this research was not
observed. The best catalyst in terms of activity was found to be the
Pt-Sn bimetallic catalyst although the reason for this improvement
is not clear and warrants further investigation. However, from the
literature the Pt-Sn catalyst has been shown to be a successful cat-
alyst for a wide variety of reactions. The improvement in activity
has been linked to both geometrical structural changes [52] and
electronic modifications [53]. The reason for the improved activity
in this work is not clear and will need to be studied in more detail.
7. Conclusions
1.9%Pd/SiO2 and 1.4%Pt/SiO2 were identified as promising cat-
alysts for the liquid phase hydrogenation of furfural. Of a series of
supports investigated (SiO2 , Al-SBA-15, ß-zeolite and TiO2 ), SiO2
was the superior support as it gave the best furfuryl alcohol selec-
tivity. Subsequent comparisons to commercial palladium catalysts
found that palladium presents good furfural conversion but was
not selective to furfuryl alcohol.
Two undesired products were formed, 2-furaldehyde diethyl
acetal and difurfuryl ether when ethanol was used as the solvent.
The production of these undesirable products was significantly
reduced or eliminated when toluene was employed as solvent.
Platinum based catalysts presented high selectivity to furfuryl
alcohol. A variety of weight loadings in the range of 0.5–2 wt% were
tested and 1%Pt/SiO2 was selected for further research of bimetallic
catalysts. From the bimetallic catalysts studied, Pt-Sn was the most
active for the hydrogenation of furfural to furfuryl alcohol.
Acknowledgements
The Earth and Natural Sciences Doctoral Studies Programme is
funded under the Programme for Research in Third-Level Institu-
tions and co-funded under the European Regional Development
Fund.
The authors would like to acknowledge the research group of
Prof Ursel Bangert for their contribution to the TEM analysis in this
publication.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.cattod.2016.06.
013.
References
[1] Council Directive (EC) 2009/28/EC of 23 (2009) on promotion of the use of
energy from renewable sources, (2009).
[2] D.J. Hayes, M.H.B. Hayes, Biofuels. Bioprod. Biorefin. 3 (2009) 500.
[3] G. Centi, P. Lanzafame, S. Perathoner, Catal. Today 167 (2011) 14.
[4] A.S. Dias, S. Lima, M. Pillinger, A.A. Valente, Catal. Lett. 114 (2007) 151.
[5] A.S. Dias, S. Lima, D. Carriazo, V. Rives, M. Pillinger, A.A. Valente, J. Catal. 244
(2006) 230.
[6] K. Dussan, B. Girisuta, D. Haverty, J.J. Leahy, M.H.B. Hayes, Bioresour. Technol.
149 (2013) 216.
[7] X.J. Shi, Y.L. Wu, H.F. Yi, G. Rui, P.P. Li, M.D. Yang, G.H. Wang, Energies 4 (2011)
669.
[8] W. Riansa-ngawong, P. Prasertsan, Carbohydr. Res. 346 (2011) 103.
201
[9] K. Dussan, B. Girisuta, M. Lopes, J.J. Leahy, M.H.B. Hayes, ChemSusChem 8
(2015) 1411.
[10] K.J. Zeitsch, The Chemistry and Technology of Furfural and Its Many
By-products: Sugar Series 13, Elsevier Science, NL, 2000.
[11] W. Huang, H. Li, B. Zhu, Y.F. feng, S. Wang, S. Zhang, Ultrason. Sonochem. 14
(2007) 67.
[12] B.M. Nagaraja, A.H. Padmasri, B.D. Raju, K.S.R. Rao, Int. J. Hydrogen Energy 36
(2011) 3417.
[13] C. Xu, L. Zheng, D. Deng, J. Liu, S. Liu, Catal. Commun. 12 (2011) 996.
[14] A.S. Dias, S. Lima, M. Pillinger, A.A. Valente, Carbohydr. Res. 341 (2006) 2946.
[15] W.B.a.T. Services, Furfural chemicals and biofuels from agriculture, A report
for the rural industries research and development corporation Sydney, (2006).
[16] S. Sitthisa, D.E. Resasco, Catal. Lett. 141 (2011) 784.
[17] Z. Xinghua, W. Tiejun, M. Longlong, W. Chuangzhi, Fuel 89 (2010) 2697.
[18] X.F. Chen, H.X. Li, H.S. Luo, M.H. Qiao, Appl. Catal. A 233 (2002) 13.
[19] M.M. Villaverde, T.F. Garetto, A.J. Marchi, Catal. Commun. 58 (2015) 6.
[20] R.V. Sharma, U. Das, R. Sammynaiken, A.K. Dalai, Appl. Catal. A 454 (2013) 127.
[21] K. Yan, A. Chen, Fuel 115 (2014) 101.
[22] S. Srivastava, N. Solanki, P. Mohanty, K.A. Shah, J.K. Parikh, A.K. Dalai, Catal.
Lett. 145 (2015) 816.
[23] S. Wei, H. Cui, J. Wang, S. Zhuo, W. Yi, L. Wang, Z. Li, Particuology 9 (2011) 69.
[24] S.P. Lee, Y.W. Chen, Ind. Eng. Chem. Res. 38 (1999) 2548.
[25] B.J. Liaw, S.J. Chiang, C.H. Tsai, Y.Z. Chen, Appl. Catal. A 284 (2005) 239.
[26] R. Rodiansono, T. Hara, N. Ichikuni, S. Shimazu, Bull. Chem. React. Eng. Catal. 9
(2014) 53.
[27] V. Vetere, A.B. Merlo, J.F. Ruggera, M.L. Casella, J. Braz. Chem. Soc. 21 (2010)
914.
[28] A.B. Merlo, V. Vetere, J.M. Ramallo-Lopez, F.G. Requejo, M.L. Casella, React.
Kinet. Mech. Catal. 104 (2011) 467.
[29] A.B. Merlo, V. Vetere, J.F. Ruggera, M.L. Casella, Catal. Commun. 10 (2009)
1665.
[30] W. Yu, Y. Tang, L. Mo, P. Chen, H. Lou, X. Zheng, Bioresour. Technol. 102 (2011)
8241.
[31] W. Yu, Y. Tang, L. Mo, P. Chen, H. Lou, X. Zheng, Catal. Commun. 13 (2011) 35.
[32] K. Fulajtárova, T. Soták, M. Hronec, I. Vávra, E. Dobročka, M. Omastová, Appl.
Catal. A 502 (2015) 78.
[33] L. Baijun, L. Lianhai, W. Bingchun, C. Tianxi, I. Katsuyoshi, Appl. Catal. A 171
(1998) 117.
[34] P. Reyes, D. Salinas, C. Campos, M. Oportus, J. Murcia, H. Rojas, G. Borda, J.L.
Garcia Fierro, Quim. Nova 33 (2010) 777.
[35] D. Liu, D. Zemlyanov, T. Wu, R.J. Lobo-Lapidus, J.A. Dumesic, J.T. Miller, C.L.
Marshall, J. Catal. 299 (2013) 336.
[36] L.R. Baker, G. Kennedy, M. Van Spronsen, A. Hervier, X. Cai, S. Chen, L.-W.
Wang, G.A. Somorjai, J. Am. Chem. Soc. 134 (2012) 14208.
[37] M.J. Taylor, L.J. Durndell, M.A. Isaacs, C.M.A. Parlett, K. Wilson, A.F. Lee, G.
Kyriakou, Appl. Catal. B 180 (2016) 580.
[38] R.V. Maligal-Ganesh, C. Xiao, T.W. Goh, L.-L. Wang, J. Gustafson, Y. Pei, Z. Qi,
D.D. Johnson, S. Zhang, F. Tao, W. Huang, ACS Catal. 6 (2016) 1754.
[39] R. Srivastava, D. Srinivas, P. Ratnasamy, Microporous Mesoporous Mater. 90
(2006) 314.
[40] D.A. Bulushev, S. Beloshapkin, J.R.H. Ross, Catal. Today 154 (2010) 7.
[41] A. Corma, M.E. Domine, S. Valencia, J. Catal. 215 (2003) 294.
[42] J. Luo, J. Yu, R.J. Gorte, E. Mahmoud, D.G. Vlachos, M.A. Smith, Catal. Sci.
Technol. 4 (2014) 3074.
[43] P. Lanzafame, D.M. Temi, S. Perathoner, G. Centi, A. Macario, A. Aloise, G.
Giordano, Catal. Today 175 (2011) 435.
[44] J. Iglesias, J.A. Melero, G. Morales, J. Moreno, Y. Segura, M. Paniagua, A.
Cambra, B. Hernandez, Catalysts 5 (2015) 1911.
[45] J.M. Rubio-Caballero, S. Saravanamurugan, P. Maireles-Torres, A. Riisager,
Catal. Today 234 (2014) 233.
[46] H. Rojas, G. Borda, D. Rosas, J.J. Martinez, P. Reyes, Dyna-Colombia 75 (2008)
115.
[47] Z. Zhang, C.U. Pittman Jr., S. Sui, J. Sun, Q. Wang, Energies 6 (2013) 1568.
[48] E.M. Wewerka, E.D. Loughran, K.L. Walters, J. Appl. Polym. Sci. 15 (1971) 1437.
[49] E. Fillion, The development of new oxabicyclic-based strategies for the stereo-
and enantioselective synthesis of azepines, in: Thiepines and Thiocines,
Polysubstituted Decalins and Related Fused Polycycles, Chemistry, Vol. PhD,
University of Toronto, Toronto, 1998, p. 364.
[50] R. Portela, V.E. Garcia-Sanchez, M. Villarroel, S.B. Rasmussen, P. Avila, Appl.
Catal. A: Gen. 510 (2016) 49.
[51] A.P. Tathod, P.L. Dhepe, Bioresour. Technol. 178 (2015) 36.
[52] F.E. López-Suárez, C.T. Carvalho-Filho, A. Bueno-López, J. Arboleda, A.
Echavarría, K.I.B. Eguiluz, G.R. Salazar-Banda, J. Int. Hydrogen Energy 40
(2015) 12674.
[53] J.H. Kim, S.M. Choi, S.H. Nam, M.H. Seo, S.H. Choi, W.B. Kim, Appl. Catal. B:
Environ. 82 (2008) 89.